Glasgow expands AI traffic-signal pilot after Pollokshaws Road bus journeys improved by up to 50%

TL;DR:

• Glasgow City Council has launched a £490,000 one-year pilot of AI-powered traffic signals along Pollokshaws Road, backed by the Scottish Government’s Bus Infrastructure Fund, after a small-scale 2024 trial with tech firm SimplifAI improved bus journey times by up to 50%.
• Separate AI deployments across the city use VivaCity computer-vision sensors to count pedestrians and approaching buses at crossings such as Paisley Road West, with the data used to balance road-user priority in real time rather than relying on historic traffic patterns.
• Loughborough University’s Transport AI Innovation Centre director Dr Haitao He told Holyrood the UK is “world-leading” in transport-AI research but “slightly behind” on moving from pilot to commercially viable deployment — a procurement-strategy gap the Chartered Institution of Highways and Transportation has formally raised with Scottish parties.

The Glasgow programme is one of the most concrete devolved-government AI deployments in UK transport and lands with Transport Scotland still lacking the equivalent of the Department for Transport’s 2025 AI action plan.

Context and Background

The technical architecture mixes two patterns: VivaCity’s edge sensors providing anonymised user counts, and SimplifAI’s signal-optimisation models running the corridor-level priority logic. The combination is what allows the system to switch from queue-clearance heuristics to a dynamic balance of pedestrian, bus and general-traffic priority. Then connectivity minister Jim Fairlie’s framing — “transformational impact … to bus prioritisation across the city” — is the explicit modal-shift policy goal: faster buses to draw passengers off private cars.

The wider research context Holyrood surfaces is the TransiT hub work on digital twins for transport decarbonisation, including Dr Joe Preece’s University of Glasgow project modelling the Queen Elizabeth Hospital catchment in Birmingham, and Loughborough’s passenger-comfort sensor research. The shared point Dr He makes — that AI demand-forecasting is what makes demand-responsive and autonomous buses commercially viable — is the part most likely to determine whether these pilots scale.

The procurement-strategy critique is the structurally significant one for UK readers. Dr He’s call for procurement as a “strategic lever to bring new technologies forward” rather than the current low-risk default echoes the UK government’s wider AI Opportunities Action Plan ambitions. The current devolved-strategy gap — Transport Scotland not yet matching the DfT’s four-objective AI plan — is a meaningful drag on Scottish AI-transport investment scaling beyond the Glasgow corridor.

Looking Forward

The 12-month Pollokshaws Road pilot will produce the data point that matters: whether 50% bus-journey improvement on a small-scale test holds at corridor scale. Watch for procurement-strategy commitments in any Transport Scotland AI strategy that emerges in response to the CIHT manifesto. UK city councils outside Scotland — Manchester, Birmingham, Leeds — are likely to read the Glasgow results as the benchmark for whether AI signal-optimisation is ready for full corridor deployment rather than test-bed novelty.